Volume 12 Number 38 14 October 2014 Pages 7413–7644
Organic & Biomolecular Chemistry www.rsc.org/obc
ISSN 1477-0520
REVIEW ARTICLE Martin G. Banwell et al. RANEY® cobalt – an underutilised reagent for the selective cleavage of C–X and N–O bonds
Organic & Biomolecular Chemistry View Article Online
REVIEW
Cite this: Org. Biomol. Chem., 2014, 12, 7433
View Journal | View Issue
RANEY® cobalt – an underutilised reagent for the selective cleavage of C–X and N–O bonds Martin G. Banwell,* Matthew T. Jones, Tristan A. Reekie, Brett D. Schwartz, Shen H. Tan and Lorenzo V. White
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
RANEY® cobalt, which was first prepared in the 1930s, is known to function effectively as a catalyst for certain chemoselective reductions. However, its utility in chemical synthesis does not seem to have been fully appreciated. This first comprehensive survey of the literature on chemical transformations involving RANEY® cobalt attempts to redress matters by, among other things, highlighting the differences between the performance of this system and its much more well-known but usually less selective congener Received 5th May 2014, Accepted 17th June 2014 DOI: 10.1039/c4ob00917g www.rsc.org/obc
RANEY® nickel. A reliable method for preparing consistently effective RANEY® cobalt is presented together with a protocol that avoids the need to use it with high pressures of dihydrogen. As such, it is hoped more attention will now be accorded to the title reagent that offers considerable promise as a powerful tool for chemical synthesis, particularly in the assembly of polycyclic frameworks through tandem reductive cyclisation processes.
A. Introduction The metal-catalysed addition of dihydrogen to both multiple and single bonds represents one of the most powerful and economically significant chemical processes known. A plethora of catalytic systems is available for this purpose and refinements of these continue to be reported in the unceasing quest to maximise chemo-, diastereo- and/or enantio-selectivity.1,2 Given the vast body of research that has been reported in the area sometimes effective techniques can “fade from view” as the fascination with new systems and protocols captures the researcher’s attention. Herein we argue this may be the case with RANEY® cobalt and suggest that it is a catalytic system that warrants re-examination by virtue of the capacity it offers for achieving, among other things, selective reductions within polyfunctionalised systems. Specifically, then, this review provides (to the best of our knowledge) the first substantial survey of the primary literature (including some patent filings) on RANEY® cobalt and its applications in chemical synthesis. The work cited herein emerged from a series of searches, conducted during 2013 and early 2014, of the chemical literature using both the SciFinder Scholar™ database and the Google Scholar™ search engine.
Research School of Chemistry, Institute of Advanced Studies, The Australian National University, Canberra, ACT 0200, Australia. E-mail: [email protected]
This journal is © The Royal Society of Chemistry 2014
B. Historical development, forms of reagent, modes of preparation and sources of dihydrogen The development of nickel-based catalysts preceded that of the cobalt-based ones.3 In 1897 the Frenchmen P. Sabatier and J. B. Senderens discovered that unsaturated organic compounds could be hydrogenated in the presence of finely divided nickel and nickel oxides and the process was greatly extended by the Russian chemist V. N. Ipatieff in the very early years of the 20th century. The broad utility of such methods was recognised by the award of the Nobel Prize to Sabatier (shared with V. Grignard) in 1912. Prompted by suggestions that occasionally “ultra catalysts” were encountered on using the Sabatier/Ipatieff protocols, from 1915 Murray Raney of Chattanooga, Tennessee conducted systematic studies on the performance of the nickel/nickel oxides catalyst systems in hydrogenation reactions as a function of the ratio of the metal to its oxides. While such experiments were to no avail (in terms of identifying “ultra catalysts”), his cognate studies on the industrial production of dihydrogen through the reaction of ferro-silicon with sodium hydroxide eventually led him to investigate the properties of the material derived from alkalibased digestion of finely ground nickel–aluminium alloys. In 1927 Raney was granted a patent for his discovery of the nickel-based catalyst that bears his name.4 In 1933 another patent that “contemplates” nickel being replaced by “iron, copper, cobalt and other catalytic materials”5 was granted.
Org. Biomol. Chem., 2014, 12, 7433–7444 | 7433
View Article Online
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Review
Raney’s description of the “more or less spongy and porous state”4 of the materials he produced led to the trademarking of these systems as “Sponge Metals Catalysts™”.6 For similar reasons they are also sometimes referred to as “skeletal” catalysts.7 While RANEY® nickel was used from the mid-1920s for effecting the hydrogenation of vegetable oils under relatively mild conditions, it was Homer Adkins at the University of Wisconsin who revealed, through a comprehensive series of studies started around 1930, that this catalyst had many other useful applications.8 Adkins also made major contributions to the preparation and categorisation of the various types of RANEY® nickel9 that now include W1 to W7 forms, details of which have been tabulated.10 Probably because of the higher cost and lower activity of RANEY® cobalt, there are far fewer studies on this catalyst compared with its nickel counterpart. Seemingly, most of the various procedures available for its preparation have been summarised.11 In addition, Reeve and Eareckson have described12 the preparation of “W7” RANEY® cobalt from a ca. 100-mesh alloy derived from 40% cobalt and 60% aluminium while Chadwell and Smith13 report a very similar protocol and note that it is somewhat easier to prepare than RANEY® nickel because removal of alkali by washing can be accomplished rather quickly (a lengthy washing procedure is required for nickel).14 A subsequent and detailed study on the preparation of RANEY® cobalt by Aller15 highlighted the need to treat finemesh cobalt/aluminium alloy powders with alkali at 15–20 °C in order to obtain highly active forms of the catalyst. Interestingly, it was noted that the use of coarser powders tended to give massive, agglomerated catalysts that did not disperse effectively, resulting in lower activity. Aller also reported that the activity of the cobalt catalyst prepared at 100 °C, even from fine-mesh powders, is exceedingly low. In contrast, the activity of the corresponding nickel catalysts increased with increasing temperatures of preparation. The so-called Aller RANEY® cobalt has been described as being more active than its W7 counterpart, at least in certain contexts.16 While at various times RANEY® cobalt has been available from commercial sources, during the course of our recent studies, some of which are detailed below, we were unable to secure this catalyst by such means. Accordingly, and after considerable investigation, we have established a protocol17 that provides material suitable for consistently achieving chemoselective reductions (at least of the types required for our purposes). A pivotal aspect to this work was the identification of a source18 of nickel-free cobalt–aluminium alloy.19 The Urushibara cobalt catalysts,1c,20 which display some reducing properties similar to RANEY® cobalt, are obtained by the reduction of cobalt salts (e.g. cobalt hexachloride) with zinc dust and leaching of the resulting precipitated and finely divided cobalt metal with aqueous sodium hydroxide, hydrochloric acid or acetic acid. Cobalt boride, which is prepared by treatment of cobalt nitrate (for example) with sodium borohydride in the presence of charcoal (as a supporting agent),21 has proved to be an excellent reagent for the reduction of
7434 | Org. Biomol. Chem., 2014, 12, 7433–7444
Organic & Biomolecular Chemistry
nitriles to amines while a mono-dispersed form prepared using a reversed micellar system is reported to have activity better than that of RANEY® cobalt in terms of its capacity to effect the hydrogenation of certain olefins.22 However, given the scope of this review and the rather limited literature available on these (Urushibara) catalysts, they will not be the subjects of any further discussion here. The alkali leaching process associated with the preparation of RANEY® cobalt from the aluminium alloy normally leaves at least some dihydrogen adsorbed on its surface with the result that freshly prepared material can sometimes be used without needing to run the desired reduction process under an atmosphere of dihydrogen. Nevertheless, the norm is to run such reactions under dihydrogen, a procedure that may become essential as the catalyst “ages” through desorption of the gas. Isopropanol can serve as a source of dihydrogen as demonstrated23 by its use in the RANEY® cobalt-catalysed transfer hydrogenolysis of certain aromatic alcohols (see section E for details).
C. (i)
Hydrogenolytic reactions Selective cleavage of carbon–halogen bonds
The capacity of the title reagent to effect the cleavage of carbon–halogen bonds, either directly or in the presence of dihydrogen, appears to have been the subject of very few studies and certainly no systematic ones. Kawashima et al. have reported24 that dihydrogen in the presence of RANEY® cobalt is able to effect the reductive dechlorination of the ringfused gem-dichlorocyclopropane 1 (Scheme 1) and does so with some levels of stereoselectivity by affording a ca. 2 : 1 mixture of the epimeric endo- and exo-forms, 2 and 3 respectively, of the corresponding mono-chloro derivative. However, RANEY® cobalt is not distinctive in this regard because similar outcomes are obtained by using W7 RANEY® nickel and Urushibara nickel B as catalysts. When PtO2 is used as catalyst then a 1 : 1 mixture of compounds 2 and 3 is obtained.
Scheme 1
(ii)
Selective cleavage of carbon–sulfur bonds
While RANEY® nickel has been used extensively for the desulfurisation of a wide range of sulfur-containing compounds,10 its cobalt counterpart has been less widely exploited for this
This journal is © The Royal Society of Chemistry 2014
View Article Online
Organic & Biomolecular Chemistry
purpose. A 1959 report by Badger and co-workers established16 that the title reagent could desulfurise dibenzothiophene (4) [and thus providing biphenyl (5)] (Scheme 2) as well as a range of other compounds, but in broad terms it was deemed “less effective” than RANEY® nickel.
Review
(a material readily derived from biomass) into lactic acid28 and for the isomerisation of erythritol (accessible by the yeastmediated fermentation of sugar or sugar alcohols such as glucose and glycerol) into threitol.29 The glycerol to lactic acid conversion is presumed to involve initial dehydrogenation of the substrate to give glyceraldehyde, dehydration of the latter to pyruvaldehyde and hydroxide-mediated conversion of this last compound to lactate.28 (iv)
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Scheme 2
One instance in which RANEY® cobalt has proven superior involves the conversion of phenanthridinethione (6) into phenanthridine (7) (Scheme 3) which proceeds in 72% yield under the indicated conditions.16 In contrast, when RANEY® nickel is used to effect the same conversion some over reduction to dihydrophenanthridine is observed and thus necessitating a subsequent dehydrogenation step.
Scheme 3
The affinity of RANEY® cobalt for organosulfur compounds has been exploited in chromatographic separations. Thus, for example, the components of a 1 : 1 w/w mixture of isoeugenol and 2,5-dimethylthiophene could be readily separated from one another by methanol elution through a mixture of freshly prepared RANEY® cobalt and clean sand. The retained thiophene was then recovered by Soxhlet extraction of the adsorbent with methanol.25 (iii)
Selective cleavage of nitrogen–oxygen bonds
In 1956 Reeve and Christian demonstrated30 that while RANEY® nickel effected the exhaustive addition of dihydrogen to oxime 8 (Scheme 4) and so generating amine 9 (20%) when the same substrate was hydrogenated in the presence of RANEY® cobalt then only C–O bond cleavage took place and so affording imine 10 (46%) as the sole reaction product. Ried and Schiller have made related observations.31
Scheme 4
During the course of work directed towards the refinement of the synthesis of an intermediate associated with the manufacture of Gemifloxacin (Factive®), a novel quinolone-based antibacterial agent, Lee and co-workers established32 that treatment of the α-cyanooxime O-methyl ether 11 (Scheme 5) with RANEY® cobalt in the presence of dihydrogen afforded
Selective cleavage of carbon–oxygen bonds
There have been few studies on the capacity of RANEY® cobalt to effect C–O bond cleavage and this is almost certainly because it is not very useful for this purpose (however, see section E). Still, the contemporary focus on the deoxygenation of biomass as a means of providing sustainable sources of fuels and other chemically valuable materials has led to extensive screening of a wide range of catalysts, including the title one, for their capacities to effect the hydrogenolysis of C–O bonds.26,27 In one instance where RANEY® cobalt has been investigated for such purposes it proved to be less useful than other catalysts.26 Nevertheless, in related studies, RANEY® cobalt was shown to be effective for converting glycerol
This journal is © The Royal Society of Chemistry 2014
Scheme 5
Org. Biomol. Chem., 2014, 12, 7433–7444 | 7435
View Article Online
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Review
Organic & Biomolecular Chemistry
the diamine 12 (unspecified yield and stereochemical outcome). This product clearly results from N–O bond cleavage as well as hydrogenation of the imine and nitrile residues (no order of events implied) within the substrate. In contrast, when the reaction was carried out with in situ Boc-protection using (t-Boc)2O then the 1°-amine derivative 13 was obtained with 83% selectivity over congener 12. The selectivity for the target compound 13 rose to 91% when a nickel/chromiumdoped form of the catalyst was used while the fully reduced product predominated when RANEY® nickel alone was employed as catalyst. As part of a program of studies directed towards the synthesis of the alkaloid limaspermidine, we observed33 (Scheme 6) that treatment of a methanolic solution of compound 14 with dihydrogen in the presence of RANEY® cobalt and p-toluenesulfonic acid (p-TsOH) mono-hydrate for short periods of time afforded both the N-hydroxylated tetracyclic indole 15 (29%) and its deoxygenated counterpart 16 (29%). In contrast, sustained reaction of substrate 14 provided compound 16 (85%) exclusively as did analogous treatment of congener 15. While a full understanding of the precise ordering of events associated with these conversions requires further investigation, it seems clear that the nitroarene residue is being converted, through N–O bond cleavage, into the corresponding N-hydroxyaniline and that this then engages in an intramolecular Schiff base condensation reaction. The conversion 15 → 16 clearly demonstrates that N-hydroxyindoles can be reduced to the equivalent indoles but whether such reductions can be extended to hydroxyamines more generally remains to be determined. Obviously the nitrile group within substrate 14 is also being reduced to the corresponding primary amine during the course of these conversions and the latter moiety then engages in an intramolecular hetero-Michael addition reaction. Related chemistry leading to the 1,5-methanoazocino[4,3-b]indole framework
of the Uleine and Strychnos alkaloids has been reported.34 Aspects of such work are presented in section H. Further discussion of the general capacity of RANEY® cobalt to effect the reduction of nitriles to 1°-amines is presented in section D(ii). It is interesting to note that when compound 14 was treated under essentially the same conditions as described above but using W7 RANEY® nickel in place of its cobalt counterpart then a mixture of compound 16 (48%) and the partially cyclised compound 17 (Fig. 1, 24%) was obtained.35,36 The latter product necessarily arises from reduction of the C–C double bond of the enone moiety associated with substrate 14 at a relatively early stage in the process and thus preventing the intramolecular hetero-Michael addition reaction required for the formation of the desired compound 16. Clearly then, RANEY® cobalt is a more chemoselective catalyst in this setting than is its nickel counterpart.
Fig. 1 Product arising from the partial reductive cyclisation of compound 14.
(v)
Selective cleavage of oxygen–oxygen bonds
A RANEY® cobalt-type catalyst obtained by digesting an alloy containing 42% cobalt and 58% aluminium with 20% aqueous sodium hydroxide at 50 °C has been reported37 to facilitate the hydrogenolytic conversion of hydroperoxide 18 into the corresponding alcohol 19 in 76% yield (Scheme 7).
Scheme 7
D. Hydrogenation reactions (i)
Scheme 6
7436 | Org. Biomol. Chem., 2014, 12, 7433–7444
Selective reduction of carbonyl groups
RANEY® cobalt can be employed for the selective reduction of the carbonyl groups within various unsaturated aldehydes and ketones. Thus, for example,38 using a reaction temperature of 100 °C and a dihydrogen pressure of ca. 9 atm this catalyst has been used to effect the reduction of citral (20, Fig. 2) to a mixture of nerol (21), geraniol (22) and citronellol (23). Under such conditions a 60% conversion of the substrate was observed after 3 h and the major products of reaction were the allylic alcohols 21 and 22. Even greater efficiencies were observed when a Urushibara-type cobalt species was used for
This journal is © The Royal Society of Chemistry 2014
View Article Online
Organic & Biomolecular Chemistry
Review
and RANEY® cobalt in the presence of ammonia at 240 °C and 272 atm of dihydrogen.47 (ii)
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Fig. 2 Citral (20) and the products arising from its reduction under various conditions – see text.
this purpose. In contrast, when RANEY® nickel was used as catalyst then citronellal (24) was the major product of reaction. Various additives have been employed in conjunction with RANEY® cobalt in an effort to achieve greater selectivities in the reduction of a range of unsaturated aldehydes. In some relatively early work in this area, Hott and Kubomatsu39 established that 2-methyl-2-pentenal was efficiently converted into the corresponding allylic alcohol (2-methyl-2-penten-1-ol) on exposure to an atmosphere of dihydrogen in the presence of a mixture of RANEY® cobalt and ferrous chloride at 55 °C and using isopropanol as solvent. The yield of product increased with increasing amounts of ferrous chloride and the rapid appearance of the blue colour characteristic of Co(II) ions suggests that the cobalt of the catalyst was being substituted by iron. The impacts of adding CoCl2, MnCl2, NiCl2, PdCl2 acetic acid and triphenylphosphine to the reaction mixture have also been investigated.40,41 As a result it has been concluded that both protic and Lewis acidic modifiers promote the formation of 2-methyl-2-penten-1-ol from the precursor aldehyde. Similarly, the hydrogenation of cinnamaldehyde to cinnamyl alcohol over RANEY® cobalt was greatly improved by modification of the catalyst with certain heteropolyacids such as Cu3/2PMo12O40.42 Related outcomes were observed in the conversion of crotonaldehyde to crotyl alcohol.43 The ketonic residue within a range of non-conjugated enones is selectively reduced by dihydrogen in the presence of W4 RANEY® cobalt and sodium hydroxide while carbon monoxide prevents such processes.44 The potentially enantioselective reduction of acetylacetone and ethyl acetoacetate by dihydrogen in the presence of RANEY® cobalt and either S,S or R,R-tartaric acid or one of eight optically active amino acids has been investigated.45 However, only modest conversions and optical yields were observed. While such observations could be considered promising, the high yielding, enantioselective and ruthenium-based CvO reduction protocols developed by the likes of Noyori have overtaken such procedures.46 The reductive mono-deoxygenation of succinimide (25) to 2-pyrrolidone (26) has been effected (Scheme 8) by dihydrogen
Selective reduction of nitriles
Undoubtedly, the reduction of nitrile groups represents the most widely exploited application of RANEY® cobalt in chemical synthesis. While there are suggestions that, in certain settings at least, RANEY® nickel represents a superior catalyst for such purposes,48 the literature is replete with examples indicating that RANEY® cobalt is highly effective. So, for instance, Reeve and Eareckson have shown12 that the reduction of 3,4methylenedioxyphenylacetonitrile to homopiperonylamine is best accomplished (88% yield) in the presence of W7 RANEY® cobalt and ammonia at 125–150 °C and using 200 atm of dihydrogen. When the same transformation is carried out using RANEY® nickel then the product amine is obtained in 82% yield. The reduction of p-cyanobenzoic acid (27) to the corresponding benzylamine 28 (Scheme 9) is readily effected at 25 °C using 3 atm of dihydrogen in aqueous ammonia and W6 or W7 RANEY® cobalt.49 Once again, ammonia is used in the reduction to suppress the formation of 2°- and 3°-amines that would otherwise be formed through reductive alkylation processes.50,51
Scheme 9
As is the case with other types of reduction reactions, the use of various additives and/or refined modes of preparation of RANEY® cobalt have been reported to enhance the efficiencies of nitrile reduction processes.52,53 Nitrile reductions can be accomplished without attendant reductive dechlorination as demonstrated by the conversion of compound 29 into the corresponding 1°-amine 30 (Scheme 10), a step that has been employed in the industrial production of certain propylamine-based fungicides.54
Scheme 10
Scheme 8
This journal is © The Royal Society of Chemistry 2014
As part of a study directed towards establishing a practical and scalable synthesis of the drug Eflornithine (used for the treatment of, among other things, African sleeping sickness),55 an ethanolic solution of nitrile 31 (Scheme 11) was treated with dihydrogen in the presence of RANEY® cobalt and
Org. Biomol. Chem., 2014, 12, 7433–7444 | 7437
View Article Online
Review
Organic & Biomolecular Chemistry
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
thereby producing the piperidone 32 “exclusively” and in 36% yield. In contrast, when the Schiff base, 33, obtained by condensing substrate 31 with benzophenone, was treated under analogous conditions and the crude reaction mixture then subjected to an aqueous work-up and treatment with HCl then compound 34 was obtained with “>99% selectivity” and in 79% yield. In neither instance was there any suggestion that competing hydrogenolysis of the C–F groups had occurred and nor was there any evidence of reduction of imine residue within compound 33.
conversions of >90% when methanol in ammonia is used as the reaction medium.58 Lower selectivities are observed with RANEY® nickel. When geranylnitrile (37, Fig. 3) was subjected to hydrogenation over RANEY® nickel doped with chromium then 3,7-dimethyl-6-octene-nitrile (38) was the favoured product of reaction while analogous treatment of the same substrate with the equivalently doped RANEY® cobalt catalyst gave the allylic amine 39 as the first-formed product of reaction. Sustained exposure of compound 39 to the reaction conditions led to the formation of mono-unsaturated congener 40.59
Fig. 3 Geranylnitrile (37) and the products arising from its reduction under various conditions – see text.
Scheme 11
During the course of the development of the synthesis of the 5-lipoxygenase inhibitor MK-0633,56 the coumarin-based nitrile 35 (Scheme 12) was successfully reduced with RANEY® cobalt in the presence of methanolic ammonia to give the 7-aminomethyl-substituted congener 36 (62%). Subsequently, the reduction was carried out using aqueous acetic acid as solvent because of the instability of the product in original reaction medium. It is notable that, (i) no hydrogenolysis of the benzylic C–N bond within the product 36 was observed and (ii) no reduction of the CvC bond within the heterocyclic ring of either the substrate or the product took place. Acetal residues also survive during the course of the RANEY® cobalt catalysed reductions of nitriles.57
More “elaborate” examples of the chemoselective reduction of nitriles embedded within polyfunctionalised substrates are presented in section H. Using chromium-doped RANEY® cobalt, polynitriles derived from the “capping”, via hetero-Michael addition reactions, of polyols or polyamines with acrylonitrile, can be reduced to the corresponding poly 1°-amines, including ones embodied in dendritic like structures.60–62 In those instances where exhaustive reduction of the nitrile groups cannot be accomplished using this catalyst system then a Urushibara-type reagent or a dissolving metal protocol appear to be effective.61 (iii)
Reduction of alkynes, alkenes and arenes
In many respects RANEY® cobalt is “prized” as a catalyst because of the capacity it offers to effect the reduction or reductive cleavage of various nitrogen- and sulfur-based functional groups in substrates also containing C–C multiple bonds. Accordingly, there is only a very modest literature dealing with the application of the title system to the reduction of alkynes, alkenes or arenes. Indeed, and perhaps a little surprisingly, we have been unable to locate a single report of the subjection of an alkyne to reaction with dihydrogen in the presence of RANEY® cobalt. The exhaustive hydrogenation of the arene thymol (41) (Scheme 13) using RANEY® cobalt and dihydrogen has been reported to deliver (±)-isomenthol (42) in 80% yield together
Scheme 12
The capacity of RANEY® cobalt to selectively effect the hydrogenation of unsaturated nitriles and thereby generating unsaturated 1°-amines is an important feature of this catalyst. Thus, for example, E-cinnamonitrile is chemoselectively reduced to 3-phenylallylamine with up to 80% selectivity and
7438 | Org. Biomol. Chem., 2014, 12, 7433–7444
Scheme 13
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Organic & Biomolecular Chemistry
with minor amounts of the racemic modifications of neomenthol, neoisomenthol and menthol.63 Various forms of RANEY® cobalt generated under differing leaching conditions and/or containing additives have been studied as catalysts for the exhaustive hydrogenation of squalene (representing a model compound for the purposes of developing protocols that could eventually allow for the “upgrading” of bio-oil produced by the highly productive algae Botryococcus braunii).64 Aller65 has investigated the effects of the promoters triethylamine and chloroplatinic acid (or combinations thereof ) on the capacity of RANEY® cobalt to reduce a range of simple substrates containing both a carbonyl group and CvC units. When substrates such as crotonaldehyde, mesityl oxide and ethyl crotonate were involved then the CvC unit was reduced chemoselectively. Such outcomes stand in stark contrast to the those detailed at section D(i) above and serve to emphasise the extraordinarily significant roles that additives/promoters can play in influencing the outcomes of reactions involving RANEY® cobalt (and, indeed, RANEY® metals more generally). Interestingly, the activity of Aller’s triethylamine–chloroplatinic acid-doped RANEY® cobalt is such that, at a temperature of 100 °C, acetophenone could be reduced to ethylbenzene, presumably via hydrogenolysis of the initially produced α-phenethyl alcohol.65 In contrast, we have observed that when a methanolic solution of allylbenzene 43 (Fig. 4) was subjected to reaction, at ca. 25 °C, with 1 atm of dihydrogen in the presence RANEY® cobalt (prepared using our protocol17) then the corresponding ethylbenzene 44 was produced in high yield and without any accompanying hydrogenolysis of the benzyl group.66
Review
observed when RANEY® cobalt was employed under the same conditions. The protocols detailed immediately above, which seem to be underutilised, avoid the need to handle gaseous dihydrogen while work up simply involves filtration and concentration of the filtrate.
F. Reactions involving (deliberate) racemisation RANEY® cobalt is known to facilitate the dehydrogenation of a range of 2°-alcohols and -amines and thereby affording the corresponding ketone or imine that can themselves accept dihydrogen under appropriate conditions and so regenerating the original alcohol or amine. When optically active substrates are subjected to such conditions then the corresponding racemates are obtained. So for example,68 when an ethereal solution of 63% ee R-(+)-2-amino-1-butanol (45, Fig. 5) was heated with RANEY® cobalt at 140 °C and 17 atm of dihydrogen for 6.75 h then the corresponding racemate was obtained in 96% yield. Resolution of this mixture of enantiomers through the formation and fractional recrystallisation of the derived tartrate salts then provided the essentially pure S-enantiomer required for the manufacture of ethambutol, a bacteriostatic antimycobacterial drug used to treat tuberculosis.
Fig. 5
Fig. 4 Allylbenzene 43 and the product, 44, of its hydrogenation in the presence of Raney® cobalt.
E.
Transfer hydrogenation reactions
RANEY® cobalt (in this case the 2700 material obtained from W. R. Grace) has been shown67 to catalyse the transfer of hydrogen from isopropanol and by such means α-substituted aromatic alcohols containing two or more aromatic rings can be deoxygenated selectively. So, for example, each of 1,2-diphenylethanol, benzhydrol, triphenylmethanol, 9-hydroxyfluorene, 1-(2-fluorenyl)ethanol, 1-(1-naphthyl)-ethanol and 1-(2-naphthyl)ethanol can be cleaved to give the corresponding aromatic hydrocarbon in quantitative yield. RANEY® nickel is less efficient in this regard but can be used to effect the hydrogenolysis of non-benzylic alcohols while its cobalt counterpart cannot. So, for example, 5-phenyl-1-pentanol is converted into pentylbenzene in 81% yield upon exposure of the former compound to RANEY® nickel and isopropanol but no reaction is
This journal is © The Royal Society of Chemistry 2014
Structure of R-(+)-2-amino-1-butanol.
R-Quinuclidin-3-ol (46, Fig. 6) represents a common pharmacophore associated with neuromodulators acting at cholinergic muscarinic receptors. Resolution of the corresponding racemate by various means provides the best way of generating this material and so methods for recycling the otherwise unusable S-enantiomer via its racemisation have been sought. One of the most effective methods for doing so involves treating the S-enantiomer with RANEY® cobalt in o-xylene under ca. 5 atm of dihydrogen and thereby generating the corresponding racemate in 97% yield.69 When the same reaction was carried out in the absence of dihydrogen then the ketone was obtained.
Fig. 6
Structure of R-Quinuclidin-3-ol.
A one-pot, dynamic kinetic resolution of certain 2°-amines has been reported wherein the unwanted enantiomer is racemised using RANEY® cobalt in the presence of dihydrogen
Org. Biomol. Chem., 2014, 12, 7433–7444 | 7439
View Article Online
Review
Organic & Biomolecular Chemistry
and the desired component (enantiomer) of the resulting mixture is then acetylated enzymatically with Candida Antarctica lipase B.70 RANEY® nickel was generally inferior in this context. Scheme 16
G.
Miscellaneous transformations
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
It has been noted71 that RANEY® cobalt allows for the (sometimes high yielding) reaction of unsaturated amides with carbon monoxide to give cyclic imides. So, for example, the cyclohexenecarbonamide 47 is converted into hexahydrophthalimide (48) by such means (Scheme 14).
The capacity of RANEY® cobalt to effect the reductive cyclisation of the α,β-unsaturated and enantiomerically pure nitrile 53 to the caprolactam 54 (Scheme 17) without significant racemisation or accompanying reduction of the CvC moiety has been reported by a Merck-based process research group.76 Best results were obtained using W. R. Grace’s Fe, Cr and Ni-doped RANEY® cobalt 2724. They noted that when RANEY® nickel was used for the same purpose then the CvC double bond was also reduced. With Rh and Pd-based systems only the CvC unit was removed (and so no cyclisation reaction occurred).
Scheme 14
In a related vein, it has recently been shown72 that RANEY® cobalt serves as an effective and recyclable catalyst in the intramolecular Pauson–Khand reaction. Treatment (Scheme 15) of the enyne 49, for example, with carbon monoxide (at pressures of ca. 23 atm) and RANEY® cobalt in THF at 130 °C for 16 h afforded the anticipated diquinane 50 in ca. 91% yield.
Scheme 17
In a true Domino process,77 and as part of a synthetic program directed toward the synthesis of the Amaryllidaceae alkaloid lycorine, we observed78 (Scheme 18) that when compound 55, embodying nitrile, alkenic and allylic C–O moieties, was subjected to reaction with ca. 30 atm of dihydrogen in Scheme 15
Somewhat more esoteric uses of RANEY® cobalt include, (i) its catalysis of the Fischer–Tropsch-like reductive oligomerisation of isonitriles in the presence of dihydrogen and liquid ammonia73 and, (ii) its conversion into dicobalt octacarbonyl under high pressures of carbon monoxide at elevated temperatures.74
H.
Domino and related reactions
The chemoselectivities displayed in RANEY® cobalt catalysed reduction processes mean that these types of transformations can deliver new functionalities in the presence of others (not be affected by the reduction) and so establishing possibilities for cyclisation reactions to take place (see Scheme 6 for example). The RANEY® cobalt-catalysed reductive cyclisation of di(cyanomethyl)amine (51) to give piperazine (52) (Scheme 16) provided an early illustration of the possibilities in this area.75
7440 | Org. Biomol. Chem., 2014, 12, 7433–7444
Scheme 18
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Organic & Biomolecular Chemistry
methanol–ammonia in the presence of RANEY® cobalt then the initially formed 1°-amine 56 (not isolated) engaged in an intramolecular SN′ reaction with the product amino-acid 57 (not isolated either) then undergoing lactamisation to give the observed product 58 (65%) that embodies the polycyclic framework of the target natural product. We have since found79 that by using larger quantities of RANEY® cobalt (normally ca. 200% w/w catalyst relative to substrate) then closely related nitriles can be reduced to the corresponding 1°-amine using just 1 atm of dihydrogen. Clearly, this makes for an operationally much simpler and thus far more attractive procedure. Interestingly, when compound 55 was subjected to attempted reduction using RANEY® nickel then only a complex mixture of products was obtained. Upon subjecting the readily accessible and polyfunctionalised cyclohexenone 59 (Scheme 19) to a range of reductive cyclisation conditions then either partially reduced and partially cyclised, fully reduced and partially cyclised or fully reduced and completely cyclised products could be obtained.34 So, treatment of the compound with dihydrogen in the presence of 10% Pd on C in methanol at 18 °C afforded indole 60 (98%) as a result of reduction of the CvC bond and reductive cyclisation of the nitro-group onto the pendant CvO (no particular order of events implied). The associated nitrile residue was clearly not affected under these conditions. On exposure to dihydrogen in the presence of RANEY® nickel and p-toluenesulfonic acid monohydrate ( p-TsOH·H2O – this acid is added to prevent reductive alkylation reactions involving the intermediates arising from nitrile group reduction) substrate 59 engaged in the same set of reactions as encountered in the first process but, in addition, the nitrile group was also reduced to the corresponding 1°-amine (and so affording compound 61 which was obtained in 87% yield). Amine 61 was also formed (in 98% yield) when compound 60 was treated with dihydrogen in the presence of RANEY® nickel. Finally,
Review
and most importantly, when cyclohexenone 59 was treated with dihydrogen, RANEY® cobalt and p-TsOH·H2O then two reductive cyclisation events took place, one involving an intramolecular hetero-Michael addition reaction (leading to the formation of a piperidine ring) and the other in which an intramolecular Schiff base type condensation reaction took place (leading to the formation of an indole moiety). As a result the tetracyclic compound 62 was obtained in 72% yield. The conversion 59 → 62 is of particular interest because the product embodies the 1,5-methanoazocino[4,3-b]indole framework associated with the Uleine and Strychnos alkaloids.
I. Effect of additives on the reactivity of RANEY® cobalt As noted on a number of occasions in the foregoing sections, the presence of additives can have a dramatic effect on the catalytic properties of Raney® cobalt.19,39–43,64,65 The manner in which the material is prepared ( particularly in terms of the initial digestion process and the subsequent washing protocols that are followed) also has an impact on activity14 and is likely the result, in part at least, of the presence of residual aluminium-based species at the surface of the reagent. Beyond the effects noted above, additional studies80 of this topic have revealed that, for example, the efficiency of reduction of various nitriles by RANEY® cobalt containing iron is strongly influenced by the presence of a third component selected from the group consisting of Ni, Rh, Ru, Pd, Pt, Os and Ir or mixtures thereof. Chromium species can also have significant effects81 as can molybdenum and tungsten.82 Similarly, when RANEY® cobalt is dispersed in an aliphatic 1°-, 2°- or 3°-amine, a diamine or a cyclic one83 then alterations in reactivity are observed. The preservation of the activity of RANEY® cobalt (and nickel) between runs through its storage in glycols has been advocated.84
J.
Scheme 19
This journal is © The Royal Society of Chemistry 2014
Comparisons with RANEY® nickel
In the foregoing sections, frequent comparisons have been made between the catalytic effects of RANEY® cobalt and its nickel counterpart. It is quite clear from these that the former species is the less reactive one. Indeed, the substantial change in the reactivity of RANEY® cobalt as a function of the presence of even small quantities (1–2%) of nickel19 in the starting alloy stands as testimony to the significant difference in behaviours of the two systems. The consequence of this is that RANEY® cobalt frequently allows for more chemoselective reactions to be carried out, most particularly when the synthesis chemist is seeking to reduce nitro- and nitrile-based moieties in the presence of otherwise potentially reactive species.
Org. Biomol. Chem., 2014, 12, 7433–7444 | 7441
View Article Online
Review
Organic & Biomolecular Chemistry
K. Mechanistic studies The literature focused on elucidating the mechanisms of reactions that employ RANEY® cobalt is rather sparse. Inelastic neutron scattering studies85 of dihydrogen and butyronitrile adsorbed on various RANEY® cobalt catalyst systems have led to the detection of partially hydrogenated and surface-associated species including ( probably) those that arise from reductive alkylation events. Intriguingly, nitrene-like intermediates appear to have been detected during the hydrogenation of CD3CN on RANEY® cobalt.86
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
L. Conclusions While this survey of a rather scattered body of literature on RANEY® cobalt reveals that it is a reagent capable of facilitating a number of distinctly fascinating chemical transformations, most notably various chemoselective reduction reactions, it is by no means one that routinely comes to mind when synthesis chemists contemplate the vast “tool set” at their disposal. There are several factors that have contributed to this situation, the first being the seeming lack of any compilation of the suite of chemical transformations that RANEY® cobalt can facilitate. Hopefully this review will help redress matters in this regard. A second factor contributing to the lack of an extended “uptake” of the title reagent in chemical synthesis is the seemingly disparate array of techniques detailed in the literature for preparing RANEY® cobalt and its “doped”/ “spiked” variants, all of which appear to display quite distinct reactivity profiles. The discontinuation of the supply of certain previously commercially available sources of RANEY® cobalt87 has certainly exacerbated matters. Furthermore, the seemingly frequent need to use very high pressures of dihydrogen (and, often simultaneously, high temperatures) is also off-putting with the end result that even for all the potential chemoselectivities “on offer” the operational requirements associated with using the reagent just seem too daunting. We believe the drawbacks mentioned above are all readily addressed by using the simple protocols defined below17 that allow for the ready preparation of RANEY® cobalt possessing consistent and effective properties. Furthermore, by using larger quantities of RANEY® cobalt (e.g. ca. 200% w/w wrt substrate)34,79b in the reduction reactions many of these can be carried out using dihydrogen at atmospheric pressure and usually at room temperature. Our recent88 application of such protocols to the tandem reductive cyclisation of the enone 63 (Scheme 20) to compound 6489 (and embodying the ABCD framework of the important alkaloid vindoline90) has served to reinforce our enthusiasm for RANEY® cobalt as a catalyst for the chemoselective reduction of polyfunctionalised substrates. We contend that RANEY® cobalt is an under utilised but particularly effective catalyst for the chemoselective reduction of polyfunctionalised organic substrates, notably those incorporating nitro and nitrile moieties. As such, continued and systematic investigations into its catalytic properties are to be
7442 | Org. Biomol. Chem., 2014, 12, 7433–7444
Scheme 20
encouraged as these are very likely to reveal new and valuable patterns of reactivity.
Notes and references 1 (a) H.-U. Blaser, C. Malan, B. Pugin, F. Spindler, H. Steiner and M. Studer, Adv. Synth. Catal., 2003, 345, 103; (b) J. G. de Vries and C. J. Elsevier, Handbook of Homogeneous Hydrogenation, Wiley-VCH, Weinheim, 2007; (c) S. Nishimura, Handbook of Heterogeneous Catalytic Hydrogenation for Organic Synthesis, John Wiley and Sons, New York, 2001. 2 The development of metal-free hydrogenation systems is an emerging area that is likely to assume considerable significance in due course. For a recent review see: L. J. Hounjet and D. W. Stephan, Org. Process Res. Dev., 2014, 18, 385. 3 For useful historical introductions see: (a) M. Raney, Ind. Eng. Chem., 1940, 32, 1199; (b) D. S. Tarbell and A. T. Tarbell, J. Chem. Educ., 1977, 54, 26. 4 M. Raney, U.S. Pat., 1,628,190, 1927. 5 M. Raney, U.S. Pat., 1,915,473, 1933. 6 See the Johnson Matthey website for information on commercially available sponge metal catalysts: http://www.jmcatalysts.com/pct/marketshome.asp?marketid=241 (accessed March 2014). 7 A. J. Smith, L. O. Garciano II, T. Tran and M. S. Wainwright, Ind. Eng. Chem. Res., 2008, 47, 1409. 8 H. Adkins, Reactions of Hydrogen with Organic Compounds Over Copper-Chromium Oxide and Nickel Catalysts, The University of Wisconsin Press, Madison, 1937. 9 (a) H. Adkins and A. A. Pavlic, J. Am. Chem. Soc., 1947, 69, 3039; (b) H. Adkins and H. R. Billica, J. Am. Chem. Soc., 1948, 70, 695; (c) H. Adkins and H. R. Billica, J. Am. Chem. Soc., 1948, 70, 3118; (d) H. Adkins and H. R. Billica, J. Am. Chem. Soc., 1948, 70, 3121 See, also: (e) R. Mozingo, Org. Synth., 1941, 21, 15; (f) H. R. Billica and H. Adkins, Org. Synth., 1949, 29, 24; (g) A. I. Vogel, A Text Book of Practical Organic Chemistry, Longmans, London, 3rd edn, 1956, pp. 870–872. 10 T.-K. Yang and D.-S. Lee, Raney Nickel, in Handbook of Reagents for Organic Synthesis: Oxidizing and Reducing Agents, ed. S. D. Burke and R. L. Danheiser, John Wiley and Sons, Chichester, UK, 1999. 11 See ref. 1c, pp. 24–25. 12 W. Reeve and W. M. Eareckson III, J. Am. Chem. Soc., 1950, 72, 3299.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Organic & Biomolecular Chemistry
13 A. J. Chadwell Jr. and H. A. Smith, J. Phys. Chem., 1956, 60, 1339. 14 The nature of the alkali-leaching process used to remove Al from the Co–Al alloy and its impact on the properties of the RANEY® cobalt so-formed has been the subject of a number of studies: (a) J. P. Orchard, A. D. Tomsett, M. S. Wainwright and D. J. Young, J. Catal., 1983, 84, 189; (b) S. Nishimura, M. Kawashima, S. Inoue, S. Takeoka, M. Shimizu and Y. Takagi, Appl. Catal., 1991, 76, 19; (c) K. Abaciouğlu and Y. Salt, React. Kinet., Mech. Catal., 2010, 101, 163. 15 B. V. Aller, J. Appl. Chem., 1957, 7, 130. 16 G. M. Badger, M. Kowanko and W. H. F. Sasse, J. Chem. Soc., 1959, 440. 17 Nickel-free cobalt–aluminium alloy (12.0 g of 3 : 7 w/w material provided as a
Organic & Biomolecular Chemistry www.rsc.org/obc
ISSN 1477-0520
REVIEW ARTICLE Martin G. Banwell et al. RANEY® cobalt – an underutilised reagent for the selective cleavage of C–X and N–O bonds
Organic & Biomolecular Chemistry View Article Online
REVIEW
Cite this: Org. Biomol. Chem., 2014, 12, 7433
View Journal | View Issue
RANEY® cobalt – an underutilised reagent for the selective cleavage of C–X and N–O bonds Martin G. Banwell,* Matthew T. Jones, Tristan A. Reekie, Brett D. Schwartz, Shen H. Tan and Lorenzo V. White
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
RANEY® cobalt, which was first prepared in the 1930s, is known to function effectively as a catalyst for certain chemoselective reductions. However, its utility in chemical synthesis does not seem to have been fully appreciated. This first comprehensive survey of the literature on chemical transformations involving RANEY® cobalt attempts to redress matters by, among other things, highlighting the differences between the performance of this system and its much more well-known but usually less selective congener Received 5th May 2014, Accepted 17th June 2014 DOI: 10.1039/c4ob00917g www.rsc.org/obc
RANEY® nickel. A reliable method for preparing consistently effective RANEY® cobalt is presented together with a protocol that avoids the need to use it with high pressures of dihydrogen. As such, it is hoped more attention will now be accorded to the title reagent that offers considerable promise as a powerful tool for chemical synthesis, particularly in the assembly of polycyclic frameworks through tandem reductive cyclisation processes.
A. Introduction The metal-catalysed addition of dihydrogen to both multiple and single bonds represents one of the most powerful and economically significant chemical processes known. A plethora of catalytic systems is available for this purpose and refinements of these continue to be reported in the unceasing quest to maximise chemo-, diastereo- and/or enantio-selectivity.1,2 Given the vast body of research that has been reported in the area sometimes effective techniques can “fade from view” as the fascination with new systems and protocols captures the researcher’s attention. Herein we argue this may be the case with RANEY® cobalt and suggest that it is a catalytic system that warrants re-examination by virtue of the capacity it offers for achieving, among other things, selective reductions within polyfunctionalised systems. Specifically, then, this review provides (to the best of our knowledge) the first substantial survey of the primary literature (including some patent filings) on RANEY® cobalt and its applications in chemical synthesis. The work cited herein emerged from a series of searches, conducted during 2013 and early 2014, of the chemical literature using both the SciFinder Scholar™ database and the Google Scholar™ search engine.
Research School of Chemistry, Institute of Advanced Studies, The Australian National University, Canberra, ACT 0200, Australia. E-mail: [email protected]
This journal is © The Royal Society of Chemistry 2014
B. Historical development, forms of reagent, modes of preparation and sources of dihydrogen The development of nickel-based catalysts preceded that of the cobalt-based ones.3 In 1897 the Frenchmen P. Sabatier and J. B. Senderens discovered that unsaturated organic compounds could be hydrogenated in the presence of finely divided nickel and nickel oxides and the process was greatly extended by the Russian chemist V. N. Ipatieff in the very early years of the 20th century. The broad utility of such methods was recognised by the award of the Nobel Prize to Sabatier (shared with V. Grignard) in 1912. Prompted by suggestions that occasionally “ultra catalysts” were encountered on using the Sabatier/Ipatieff protocols, from 1915 Murray Raney of Chattanooga, Tennessee conducted systematic studies on the performance of the nickel/nickel oxides catalyst systems in hydrogenation reactions as a function of the ratio of the metal to its oxides. While such experiments were to no avail (in terms of identifying “ultra catalysts”), his cognate studies on the industrial production of dihydrogen through the reaction of ferro-silicon with sodium hydroxide eventually led him to investigate the properties of the material derived from alkalibased digestion of finely ground nickel–aluminium alloys. In 1927 Raney was granted a patent for his discovery of the nickel-based catalyst that bears his name.4 In 1933 another patent that “contemplates” nickel being replaced by “iron, copper, cobalt and other catalytic materials”5 was granted.
Org. Biomol. Chem., 2014, 12, 7433–7444 | 7433
View Article Online
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Review
Raney’s description of the “more or less spongy and porous state”4 of the materials he produced led to the trademarking of these systems as “Sponge Metals Catalysts™”.6 For similar reasons they are also sometimes referred to as “skeletal” catalysts.7 While RANEY® nickel was used from the mid-1920s for effecting the hydrogenation of vegetable oils under relatively mild conditions, it was Homer Adkins at the University of Wisconsin who revealed, through a comprehensive series of studies started around 1930, that this catalyst had many other useful applications.8 Adkins also made major contributions to the preparation and categorisation of the various types of RANEY® nickel9 that now include W1 to W7 forms, details of which have been tabulated.10 Probably because of the higher cost and lower activity of RANEY® cobalt, there are far fewer studies on this catalyst compared with its nickel counterpart. Seemingly, most of the various procedures available for its preparation have been summarised.11 In addition, Reeve and Eareckson have described12 the preparation of “W7” RANEY® cobalt from a ca. 100-mesh alloy derived from 40% cobalt and 60% aluminium while Chadwell and Smith13 report a very similar protocol and note that it is somewhat easier to prepare than RANEY® nickel because removal of alkali by washing can be accomplished rather quickly (a lengthy washing procedure is required for nickel).14 A subsequent and detailed study on the preparation of RANEY® cobalt by Aller15 highlighted the need to treat finemesh cobalt/aluminium alloy powders with alkali at 15–20 °C in order to obtain highly active forms of the catalyst. Interestingly, it was noted that the use of coarser powders tended to give massive, agglomerated catalysts that did not disperse effectively, resulting in lower activity. Aller also reported that the activity of the cobalt catalyst prepared at 100 °C, even from fine-mesh powders, is exceedingly low. In contrast, the activity of the corresponding nickel catalysts increased with increasing temperatures of preparation. The so-called Aller RANEY® cobalt has been described as being more active than its W7 counterpart, at least in certain contexts.16 While at various times RANEY® cobalt has been available from commercial sources, during the course of our recent studies, some of which are detailed below, we were unable to secure this catalyst by such means. Accordingly, and after considerable investigation, we have established a protocol17 that provides material suitable for consistently achieving chemoselective reductions (at least of the types required for our purposes). A pivotal aspect to this work was the identification of a source18 of nickel-free cobalt–aluminium alloy.19 The Urushibara cobalt catalysts,1c,20 which display some reducing properties similar to RANEY® cobalt, are obtained by the reduction of cobalt salts (e.g. cobalt hexachloride) with zinc dust and leaching of the resulting precipitated and finely divided cobalt metal with aqueous sodium hydroxide, hydrochloric acid or acetic acid. Cobalt boride, which is prepared by treatment of cobalt nitrate (for example) with sodium borohydride in the presence of charcoal (as a supporting agent),21 has proved to be an excellent reagent for the reduction of
7434 | Org. Biomol. Chem., 2014, 12, 7433–7444
Organic & Biomolecular Chemistry
nitriles to amines while a mono-dispersed form prepared using a reversed micellar system is reported to have activity better than that of RANEY® cobalt in terms of its capacity to effect the hydrogenation of certain olefins.22 However, given the scope of this review and the rather limited literature available on these (Urushibara) catalysts, they will not be the subjects of any further discussion here. The alkali leaching process associated with the preparation of RANEY® cobalt from the aluminium alloy normally leaves at least some dihydrogen adsorbed on its surface with the result that freshly prepared material can sometimes be used without needing to run the desired reduction process under an atmosphere of dihydrogen. Nevertheless, the norm is to run such reactions under dihydrogen, a procedure that may become essential as the catalyst “ages” through desorption of the gas. Isopropanol can serve as a source of dihydrogen as demonstrated23 by its use in the RANEY® cobalt-catalysed transfer hydrogenolysis of certain aromatic alcohols (see section E for details).
C. (i)
Hydrogenolytic reactions Selective cleavage of carbon–halogen bonds
The capacity of the title reagent to effect the cleavage of carbon–halogen bonds, either directly or in the presence of dihydrogen, appears to have been the subject of very few studies and certainly no systematic ones. Kawashima et al. have reported24 that dihydrogen in the presence of RANEY® cobalt is able to effect the reductive dechlorination of the ringfused gem-dichlorocyclopropane 1 (Scheme 1) and does so with some levels of stereoselectivity by affording a ca. 2 : 1 mixture of the epimeric endo- and exo-forms, 2 and 3 respectively, of the corresponding mono-chloro derivative. However, RANEY® cobalt is not distinctive in this regard because similar outcomes are obtained by using W7 RANEY® nickel and Urushibara nickel B as catalysts. When PtO2 is used as catalyst then a 1 : 1 mixture of compounds 2 and 3 is obtained.
Scheme 1
(ii)
Selective cleavage of carbon–sulfur bonds
While RANEY® nickel has been used extensively for the desulfurisation of a wide range of sulfur-containing compounds,10 its cobalt counterpart has been less widely exploited for this
This journal is © The Royal Society of Chemistry 2014
View Article Online
Organic & Biomolecular Chemistry
purpose. A 1959 report by Badger and co-workers established16 that the title reagent could desulfurise dibenzothiophene (4) [and thus providing biphenyl (5)] (Scheme 2) as well as a range of other compounds, but in broad terms it was deemed “less effective” than RANEY® nickel.
Review
(a material readily derived from biomass) into lactic acid28 and for the isomerisation of erythritol (accessible by the yeastmediated fermentation of sugar or sugar alcohols such as glucose and glycerol) into threitol.29 The glycerol to lactic acid conversion is presumed to involve initial dehydrogenation of the substrate to give glyceraldehyde, dehydration of the latter to pyruvaldehyde and hydroxide-mediated conversion of this last compound to lactate.28 (iv)
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Scheme 2
One instance in which RANEY® cobalt has proven superior involves the conversion of phenanthridinethione (6) into phenanthridine (7) (Scheme 3) which proceeds in 72% yield under the indicated conditions.16 In contrast, when RANEY® nickel is used to effect the same conversion some over reduction to dihydrophenanthridine is observed and thus necessitating a subsequent dehydrogenation step.
Scheme 3
The affinity of RANEY® cobalt for organosulfur compounds has been exploited in chromatographic separations. Thus, for example, the components of a 1 : 1 w/w mixture of isoeugenol and 2,5-dimethylthiophene could be readily separated from one another by methanol elution through a mixture of freshly prepared RANEY® cobalt and clean sand. The retained thiophene was then recovered by Soxhlet extraction of the adsorbent with methanol.25 (iii)
Selective cleavage of nitrogen–oxygen bonds
In 1956 Reeve and Christian demonstrated30 that while RANEY® nickel effected the exhaustive addition of dihydrogen to oxime 8 (Scheme 4) and so generating amine 9 (20%) when the same substrate was hydrogenated in the presence of RANEY® cobalt then only C–O bond cleavage took place and so affording imine 10 (46%) as the sole reaction product. Ried and Schiller have made related observations.31
Scheme 4
During the course of work directed towards the refinement of the synthesis of an intermediate associated with the manufacture of Gemifloxacin (Factive®), a novel quinolone-based antibacterial agent, Lee and co-workers established32 that treatment of the α-cyanooxime O-methyl ether 11 (Scheme 5) with RANEY® cobalt in the presence of dihydrogen afforded
Selective cleavage of carbon–oxygen bonds
There have been few studies on the capacity of RANEY® cobalt to effect C–O bond cleavage and this is almost certainly because it is not very useful for this purpose (however, see section E). Still, the contemporary focus on the deoxygenation of biomass as a means of providing sustainable sources of fuels and other chemically valuable materials has led to extensive screening of a wide range of catalysts, including the title one, for their capacities to effect the hydrogenolysis of C–O bonds.26,27 In one instance where RANEY® cobalt has been investigated for such purposes it proved to be less useful than other catalysts.26 Nevertheless, in related studies, RANEY® cobalt was shown to be effective for converting glycerol
This journal is © The Royal Society of Chemistry 2014
Scheme 5
Org. Biomol. Chem., 2014, 12, 7433–7444 | 7435
View Article Online
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Review
Organic & Biomolecular Chemistry
the diamine 12 (unspecified yield and stereochemical outcome). This product clearly results from N–O bond cleavage as well as hydrogenation of the imine and nitrile residues (no order of events implied) within the substrate. In contrast, when the reaction was carried out with in situ Boc-protection using (t-Boc)2O then the 1°-amine derivative 13 was obtained with 83% selectivity over congener 12. The selectivity for the target compound 13 rose to 91% when a nickel/chromiumdoped form of the catalyst was used while the fully reduced product predominated when RANEY® nickel alone was employed as catalyst. As part of a program of studies directed towards the synthesis of the alkaloid limaspermidine, we observed33 (Scheme 6) that treatment of a methanolic solution of compound 14 with dihydrogen in the presence of RANEY® cobalt and p-toluenesulfonic acid (p-TsOH) mono-hydrate for short periods of time afforded both the N-hydroxylated tetracyclic indole 15 (29%) and its deoxygenated counterpart 16 (29%). In contrast, sustained reaction of substrate 14 provided compound 16 (85%) exclusively as did analogous treatment of congener 15. While a full understanding of the precise ordering of events associated with these conversions requires further investigation, it seems clear that the nitroarene residue is being converted, through N–O bond cleavage, into the corresponding N-hydroxyaniline and that this then engages in an intramolecular Schiff base condensation reaction. The conversion 15 → 16 clearly demonstrates that N-hydroxyindoles can be reduced to the equivalent indoles but whether such reductions can be extended to hydroxyamines more generally remains to be determined. Obviously the nitrile group within substrate 14 is also being reduced to the corresponding primary amine during the course of these conversions and the latter moiety then engages in an intramolecular hetero-Michael addition reaction. Related chemistry leading to the 1,5-methanoazocino[4,3-b]indole framework
of the Uleine and Strychnos alkaloids has been reported.34 Aspects of such work are presented in section H. Further discussion of the general capacity of RANEY® cobalt to effect the reduction of nitriles to 1°-amines is presented in section D(ii). It is interesting to note that when compound 14 was treated under essentially the same conditions as described above but using W7 RANEY® nickel in place of its cobalt counterpart then a mixture of compound 16 (48%) and the partially cyclised compound 17 (Fig. 1, 24%) was obtained.35,36 The latter product necessarily arises from reduction of the C–C double bond of the enone moiety associated with substrate 14 at a relatively early stage in the process and thus preventing the intramolecular hetero-Michael addition reaction required for the formation of the desired compound 16. Clearly then, RANEY® cobalt is a more chemoselective catalyst in this setting than is its nickel counterpart.
Fig. 1 Product arising from the partial reductive cyclisation of compound 14.
(v)
Selective cleavage of oxygen–oxygen bonds
A RANEY® cobalt-type catalyst obtained by digesting an alloy containing 42% cobalt and 58% aluminium with 20% aqueous sodium hydroxide at 50 °C has been reported37 to facilitate the hydrogenolytic conversion of hydroperoxide 18 into the corresponding alcohol 19 in 76% yield (Scheme 7).
Scheme 7
D. Hydrogenation reactions (i)
Scheme 6
7436 | Org. Biomol. Chem., 2014, 12, 7433–7444
Selective reduction of carbonyl groups
RANEY® cobalt can be employed for the selective reduction of the carbonyl groups within various unsaturated aldehydes and ketones. Thus, for example,38 using a reaction temperature of 100 °C and a dihydrogen pressure of ca. 9 atm this catalyst has been used to effect the reduction of citral (20, Fig. 2) to a mixture of nerol (21), geraniol (22) and citronellol (23). Under such conditions a 60% conversion of the substrate was observed after 3 h and the major products of reaction were the allylic alcohols 21 and 22. Even greater efficiencies were observed when a Urushibara-type cobalt species was used for
This journal is © The Royal Society of Chemistry 2014
View Article Online
Organic & Biomolecular Chemistry
Review
and RANEY® cobalt in the presence of ammonia at 240 °C and 272 atm of dihydrogen.47 (ii)
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Fig. 2 Citral (20) and the products arising from its reduction under various conditions – see text.
this purpose. In contrast, when RANEY® nickel was used as catalyst then citronellal (24) was the major product of reaction. Various additives have been employed in conjunction with RANEY® cobalt in an effort to achieve greater selectivities in the reduction of a range of unsaturated aldehydes. In some relatively early work in this area, Hott and Kubomatsu39 established that 2-methyl-2-pentenal was efficiently converted into the corresponding allylic alcohol (2-methyl-2-penten-1-ol) on exposure to an atmosphere of dihydrogen in the presence of a mixture of RANEY® cobalt and ferrous chloride at 55 °C and using isopropanol as solvent. The yield of product increased with increasing amounts of ferrous chloride and the rapid appearance of the blue colour characteristic of Co(II) ions suggests that the cobalt of the catalyst was being substituted by iron. The impacts of adding CoCl2, MnCl2, NiCl2, PdCl2 acetic acid and triphenylphosphine to the reaction mixture have also been investigated.40,41 As a result it has been concluded that both protic and Lewis acidic modifiers promote the formation of 2-methyl-2-penten-1-ol from the precursor aldehyde. Similarly, the hydrogenation of cinnamaldehyde to cinnamyl alcohol over RANEY® cobalt was greatly improved by modification of the catalyst with certain heteropolyacids such as Cu3/2PMo12O40.42 Related outcomes were observed in the conversion of crotonaldehyde to crotyl alcohol.43 The ketonic residue within a range of non-conjugated enones is selectively reduced by dihydrogen in the presence of W4 RANEY® cobalt and sodium hydroxide while carbon monoxide prevents such processes.44 The potentially enantioselective reduction of acetylacetone and ethyl acetoacetate by dihydrogen in the presence of RANEY® cobalt and either S,S or R,R-tartaric acid or one of eight optically active amino acids has been investigated.45 However, only modest conversions and optical yields were observed. While such observations could be considered promising, the high yielding, enantioselective and ruthenium-based CvO reduction protocols developed by the likes of Noyori have overtaken such procedures.46 The reductive mono-deoxygenation of succinimide (25) to 2-pyrrolidone (26) has been effected (Scheme 8) by dihydrogen
Selective reduction of nitriles
Undoubtedly, the reduction of nitrile groups represents the most widely exploited application of RANEY® cobalt in chemical synthesis. While there are suggestions that, in certain settings at least, RANEY® nickel represents a superior catalyst for such purposes,48 the literature is replete with examples indicating that RANEY® cobalt is highly effective. So, for instance, Reeve and Eareckson have shown12 that the reduction of 3,4methylenedioxyphenylacetonitrile to homopiperonylamine is best accomplished (88% yield) in the presence of W7 RANEY® cobalt and ammonia at 125–150 °C and using 200 atm of dihydrogen. When the same transformation is carried out using RANEY® nickel then the product amine is obtained in 82% yield. The reduction of p-cyanobenzoic acid (27) to the corresponding benzylamine 28 (Scheme 9) is readily effected at 25 °C using 3 atm of dihydrogen in aqueous ammonia and W6 or W7 RANEY® cobalt.49 Once again, ammonia is used in the reduction to suppress the formation of 2°- and 3°-amines that would otherwise be formed through reductive alkylation processes.50,51
Scheme 9
As is the case with other types of reduction reactions, the use of various additives and/or refined modes of preparation of RANEY® cobalt have been reported to enhance the efficiencies of nitrile reduction processes.52,53 Nitrile reductions can be accomplished without attendant reductive dechlorination as demonstrated by the conversion of compound 29 into the corresponding 1°-amine 30 (Scheme 10), a step that has been employed in the industrial production of certain propylamine-based fungicides.54
Scheme 10
Scheme 8
This journal is © The Royal Society of Chemistry 2014
As part of a study directed towards establishing a practical and scalable synthesis of the drug Eflornithine (used for the treatment of, among other things, African sleeping sickness),55 an ethanolic solution of nitrile 31 (Scheme 11) was treated with dihydrogen in the presence of RANEY® cobalt and
Org. Biomol. Chem., 2014, 12, 7433–7444 | 7437
View Article Online
Review
Organic & Biomolecular Chemistry
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
thereby producing the piperidone 32 “exclusively” and in 36% yield. In contrast, when the Schiff base, 33, obtained by condensing substrate 31 with benzophenone, was treated under analogous conditions and the crude reaction mixture then subjected to an aqueous work-up and treatment with HCl then compound 34 was obtained with “>99% selectivity” and in 79% yield. In neither instance was there any suggestion that competing hydrogenolysis of the C–F groups had occurred and nor was there any evidence of reduction of imine residue within compound 33.
conversions of >90% when methanol in ammonia is used as the reaction medium.58 Lower selectivities are observed with RANEY® nickel. When geranylnitrile (37, Fig. 3) was subjected to hydrogenation over RANEY® nickel doped with chromium then 3,7-dimethyl-6-octene-nitrile (38) was the favoured product of reaction while analogous treatment of the same substrate with the equivalently doped RANEY® cobalt catalyst gave the allylic amine 39 as the first-formed product of reaction. Sustained exposure of compound 39 to the reaction conditions led to the formation of mono-unsaturated congener 40.59
Fig. 3 Geranylnitrile (37) and the products arising from its reduction under various conditions – see text.
Scheme 11
During the course of the development of the synthesis of the 5-lipoxygenase inhibitor MK-0633,56 the coumarin-based nitrile 35 (Scheme 12) was successfully reduced with RANEY® cobalt in the presence of methanolic ammonia to give the 7-aminomethyl-substituted congener 36 (62%). Subsequently, the reduction was carried out using aqueous acetic acid as solvent because of the instability of the product in original reaction medium. It is notable that, (i) no hydrogenolysis of the benzylic C–N bond within the product 36 was observed and (ii) no reduction of the CvC bond within the heterocyclic ring of either the substrate or the product took place. Acetal residues also survive during the course of the RANEY® cobalt catalysed reductions of nitriles.57
More “elaborate” examples of the chemoselective reduction of nitriles embedded within polyfunctionalised substrates are presented in section H. Using chromium-doped RANEY® cobalt, polynitriles derived from the “capping”, via hetero-Michael addition reactions, of polyols or polyamines with acrylonitrile, can be reduced to the corresponding poly 1°-amines, including ones embodied in dendritic like structures.60–62 In those instances where exhaustive reduction of the nitrile groups cannot be accomplished using this catalyst system then a Urushibara-type reagent or a dissolving metal protocol appear to be effective.61 (iii)
Reduction of alkynes, alkenes and arenes
In many respects RANEY® cobalt is “prized” as a catalyst because of the capacity it offers to effect the reduction or reductive cleavage of various nitrogen- and sulfur-based functional groups in substrates also containing C–C multiple bonds. Accordingly, there is only a very modest literature dealing with the application of the title system to the reduction of alkynes, alkenes or arenes. Indeed, and perhaps a little surprisingly, we have been unable to locate a single report of the subjection of an alkyne to reaction with dihydrogen in the presence of RANEY® cobalt. The exhaustive hydrogenation of the arene thymol (41) (Scheme 13) using RANEY® cobalt and dihydrogen has been reported to deliver (±)-isomenthol (42) in 80% yield together
Scheme 12
The capacity of RANEY® cobalt to selectively effect the hydrogenation of unsaturated nitriles and thereby generating unsaturated 1°-amines is an important feature of this catalyst. Thus, for example, E-cinnamonitrile is chemoselectively reduced to 3-phenylallylamine with up to 80% selectivity and
7438 | Org. Biomol. Chem., 2014, 12, 7433–7444
Scheme 13
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Organic & Biomolecular Chemistry
with minor amounts of the racemic modifications of neomenthol, neoisomenthol and menthol.63 Various forms of RANEY® cobalt generated under differing leaching conditions and/or containing additives have been studied as catalysts for the exhaustive hydrogenation of squalene (representing a model compound for the purposes of developing protocols that could eventually allow for the “upgrading” of bio-oil produced by the highly productive algae Botryococcus braunii).64 Aller65 has investigated the effects of the promoters triethylamine and chloroplatinic acid (or combinations thereof ) on the capacity of RANEY® cobalt to reduce a range of simple substrates containing both a carbonyl group and CvC units. When substrates such as crotonaldehyde, mesityl oxide and ethyl crotonate were involved then the CvC unit was reduced chemoselectively. Such outcomes stand in stark contrast to the those detailed at section D(i) above and serve to emphasise the extraordinarily significant roles that additives/promoters can play in influencing the outcomes of reactions involving RANEY® cobalt (and, indeed, RANEY® metals more generally). Interestingly, the activity of Aller’s triethylamine–chloroplatinic acid-doped RANEY® cobalt is such that, at a temperature of 100 °C, acetophenone could be reduced to ethylbenzene, presumably via hydrogenolysis of the initially produced α-phenethyl alcohol.65 In contrast, we have observed that when a methanolic solution of allylbenzene 43 (Fig. 4) was subjected to reaction, at ca. 25 °C, with 1 atm of dihydrogen in the presence RANEY® cobalt (prepared using our protocol17) then the corresponding ethylbenzene 44 was produced in high yield and without any accompanying hydrogenolysis of the benzyl group.66
Review
observed when RANEY® cobalt was employed under the same conditions. The protocols detailed immediately above, which seem to be underutilised, avoid the need to handle gaseous dihydrogen while work up simply involves filtration and concentration of the filtrate.
F. Reactions involving (deliberate) racemisation RANEY® cobalt is known to facilitate the dehydrogenation of a range of 2°-alcohols and -amines and thereby affording the corresponding ketone or imine that can themselves accept dihydrogen under appropriate conditions and so regenerating the original alcohol or amine. When optically active substrates are subjected to such conditions then the corresponding racemates are obtained. So for example,68 when an ethereal solution of 63% ee R-(+)-2-amino-1-butanol (45, Fig. 5) was heated with RANEY® cobalt at 140 °C and 17 atm of dihydrogen for 6.75 h then the corresponding racemate was obtained in 96% yield. Resolution of this mixture of enantiomers through the formation and fractional recrystallisation of the derived tartrate salts then provided the essentially pure S-enantiomer required for the manufacture of ethambutol, a bacteriostatic antimycobacterial drug used to treat tuberculosis.
Fig. 5
Fig. 4 Allylbenzene 43 and the product, 44, of its hydrogenation in the presence of Raney® cobalt.
E.
Transfer hydrogenation reactions
RANEY® cobalt (in this case the 2700 material obtained from W. R. Grace) has been shown67 to catalyse the transfer of hydrogen from isopropanol and by such means α-substituted aromatic alcohols containing two or more aromatic rings can be deoxygenated selectively. So, for example, each of 1,2-diphenylethanol, benzhydrol, triphenylmethanol, 9-hydroxyfluorene, 1-(2-fluorenyl)ethanol, 1-(1-naphthyl)-ethanol and 1-(2-naphthyl)ethanol can be cleaved to give the corresponding aromatic hydrocarbon in quantitative yield. RANEY® nickel is less efficient in this regard but can be used to effect the hydrogenolysis of non-benzylic alcohols while its cobalt counterpart cannot. So, for example, 5-phenyl-1-pentanol is converted into pentylbenzene in 81% yield upon exposure of the former compound to RANEY® nickel and isopropanol but no reaction is
This journal is © The Royal Society of Chemistry 2014
Structure of R-(+)-2-amino-1-butanol.
R-Quinuclidin-3-ol (46, Fig. 6) represents a common pharmacophore associated with neuromodulators acting at cholinergic muscarinic receptors. Resolution of the corresponding racemate by various means provides the best way of generating this material and so methods for recycling the otherwise unusable S-enantiomer via its racemisation have been sought. One of the most effective methods for doing so involves treating the S-enantiomer with RANEY® cobalt in o-xylene under ca. 5 atm of dihydrogen and thereby generating the corresponding racemate in 97% yield.69 When the same reaction was carried out in the absence of dihydrogen then the ketone was obtained.
Fig. 6
Structure of R-Quinuclidin-3-ol.
A one-pot, dynamic kinetic resolution of certain 2°-amines has been reported wherein the unwanted enantiomer is racemised using RANEY® cobalt in the presence of dihydrogen
Org. Biomol. Chem., 2014, 12, 7433–7444 | 7439
View Article Online
Review
Organic & Biomolecular Chemistry
and the desired component (enantiomer) of the resulting mixture is then acetylated enzymatically with Candida Antarctica lipase B.70 RANEY® nickel was generally inferior in this context. Scheme 16
G.
Miscellaneous transformations
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
It has been noted71 that RANEY® cobalt allows for the (sometimes high yielding) reaction of unsaturated amides with carbon monoxide to give cyclic imides. So, for example, the cyclohexenecarbonamide 47 is converted into hexahydrophthalimide (48) by such means (Scheme 14).
The capacity of RANEY® cobalt to effect the reductive cyclisation of the α,β-unsaturated and enantiomerically pure nitrile 53 to the caprolactam 54 (Scheme 17) without significant racemisation or accompanying reduction of the CvC moiety has been reported by a Merck-based process research group.76 Best results were obtained using W. R. Grace’s Fe, Cr and Ni-doped RANEY® cobalt 2724. They noted that when RANEY® nickel was used for the same purpose then the CvC double bond was also reduced. With Rh and Pd-based systems only the CvC unit was removed (and so no cyclisation reaction occurred).
Scheme 14
In a related vein, it has recently been shown72 that RANEY® cobalt serves as an effective and recyclable catalyst in the intramolecular Pauson–Khand reaction. Treatment (Scheme 15) of the enyne 49, for example, with carbon monoxide (at pressures of ca. 23 atm) and RANEY® cobalt in THF at 130 °C for 16 h afforded the anticipated diquinane 50 in ca. 91% yield.
Scheme 17
In a true Domino process,77 and as part of a synthetic program directed toward the synthesis of the Amaryllidaceae alkaloid lycorine, we observed78 (Scheme 18) that when compound 55, embodying nitrile, alkenic and allylic C–O moieties, was subjected to reaction with ca. 30 atm of dihydrogen in Scheme 15
Somewhat more esoteric uses of RANEY® cobalt include, (i) its catalysis of the Fischer–Tropsch-like reductive oligomerisation of isonitriles in the presence of dihydrogen and liquid ammonia73 and, (ii) its conversion into dicobalt octacarbonyl under high pressures of carbon monoxide at elevated temperatures.74
H.
Domino and related reactions
The chemoselectivities displayed in RANEY® cobalt catalysed reduction processes mean that these types of transformations can deliver new functionalities in the presence of others (not be affected by the reduction) and so establishing possibilities for cyclisation reactions to take place (see Scheme 6 for example). The RANEY® cobalt-catalysed reductive cyclisation of di(cyanomethyl)amine (51) to give piperazine (52) (Scheme 16) provided an early illustration of the possibilities in this area.75
7440 | Org. Biomol. Chem., 2014, 12, 7433–7444
Scheme 18
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Organic & Biomolecular Chemistry
methanol–ammonia in the presence of RANEY® cobalt then the initially formed 1°-amine 56 (not isolated) engaged in an intramolecular SN′ reaction with the product amino-acid 57 (not isolated either) then undergoing lactamisation to give the observed product 58 (65%) that embodies the polycyclic framework of the target natural product. We have since found79 that by using larger quantities of RANEY® cobalt (normally ca. 200% w/w catalyst relative to substrate) then closely related nitriles can be reduced to the corresponding 1°-amine using just 1 atm of dihydrogen. Clearly, this makes for an operationally much simpler and thus far more attractive procedure. Interestingly, when compound 55 was subjected to attempted reduction using RANEY® nickel then only a complex mixture of products was obtained. Upon subjecting the readily accessible and polyfunctionalised cyclohexenone 59 (Scheme 19) to a range of reductive cyclisation conditions then either partially reduced and partially cyclised, fully reduced and partially cyclised or fully reduced and completely cyclised products could be obtained.34 So, treatment of the compound with dihydrogen in the presence of 10% Pd on C in methanol at 18 °C afforded indole 60 (98%) as a result of reduction of the CvC bond and reductive cyclisation of the nitro-group onto the pendant CvO (no particular order of events implied). The associated nitrile residue was clearly not affected under these conditions. On exposure to dihydrogen in the presence of RANEY® nickel and p-toluenesulfonic acid monohydrate ( p-TsOH·H2O – this acid is added to prevent reductive alkylation reactions involving the intermediates arising from nitrile group reduction) substrate 59 engaged in the same set of reactions as encountered in the first process but, in addition, the nitrile group was also reduced to the corresponding 1°-amine (and so affording compound 61 which was obtained in 87% yield). Amine 61 was also formed (in 98% yield) when compound 60 was treated with dihydrogen in the presence of RANEY® nickel. Finally,
Review
and most importantly, when cyclohexenone 59 was treated with dihydrogen, RANEY® cobalt and p-TsOH·H2O then two reductive cyclisation events took place, one involving an intramolecular hetero-Michael addition reaction (leading to the formation of a piperidine ring) and the other in which an intramolecular Schiff base type condensation reaction took place (leading to the formation of an indole moiety). As a result the tetracyclic compound 62 was obtained in 72% yield. The conversion 59 → 62 is of particular interest because the product embodies the 1,5-methanoazocino[4,3-b]indole framework associated with the Uleine and Strychnos alkaloids.
I. Effect of additives on the reactivity of RANEY® cobalt As noted on a number of occasions in the foregoing sections, the presence of additives can have a dramatic effect on the catalytic properties of Raney® cobalt.19,39–43,64,65 The manner in which the material is prepared ( particularly in terms of the initial digestion process and the subsequent washing protocols that are followed) also has an impact on activity14 and is likely the result, in part at least, of the presence of residual aluminium-based species at the surface of the reagent. Beyond the effects noted above, additional studies80 of this topic have revealed that, for example, the efficiency of reduction of various nitriles by RANEY® cobalt containing iron is strongly influenced by the presence of a third component selected from the group consisting of Ni, Rh, Ru, Pd, Pt, Os and Ir or mixtures thereof. Chromium species can also have significant effects81 as can molybdenum and tungsten.82 Similarly, when RANEY® cobalt is dispersed in an aliphatic 1°-, 2°- or 3°-amine, a diamine or a cyclic one83 then alterations in reactivity are observed. The preservation of the activity of RANEY® cobalt (and nickel) between runs through its storage in glycols has been advocated.84
J.
Scheme 19
This journal is © The Royal Society of Chemistry 2014
Comparisons with RANEY® nickel
In the foregoing sections, frequent comparisons have been made between the catalytic effects of RANEY® cobalt and its nickel counterpart. It is quite clear from these that the former species is the less reactive one. Indeed, the substantial change in the reactivity of RANEY® cobalt as a function of the presence of even small quantities (1–2%) of nickel19 in the starting alloy stands as testimony to the significant difference in behaviours of the two systems. The consequence of this is that RANEY® cobalt frequently allows for more chemoselective reactions to be carried out, most particularly when the synthesis chemist is seeking to reduce nitro- and nitrile-based moieties in the presence of otherwise potentially reactive species.
Org. Biomol. Chem., 2014, 12, 7433–7444 | 7441
View Article Online
Review
Organic & Biomolecular Chemistry
K. Mechanistic studies The literature focused on elucidating the mechanisms of reactions that employ RANEY® cobalt is rather sparse. Inelastic neutron scattering studies85 of dihydrogen and butyronitrile adsorbed on various RANEY® cobalt catalyst systems have led to the detection of partially hydrogenated and surface-associated species including ( probably) those that arise from reductive alkylation events. Intriguingly, nitrene-like intermediates appear to have been detected during the hydrogenation of CD3CN on RANEY® cobalt.86
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
L. Conclusions While this survey of a rather scattered body of literature on RANEY® cobalt reveals that it is a reagent capable of facilitating a number of distinctly fascinating chemical transformations, most notably various chemoselective reduction reactions, it is by no means one that routinely comes to mind when synthesis chemists contemplate the vast “tool set” at their disposal. There are several factors that have contributed to this situation, the first being the seeming lack of any compilation of the suite of chemical transformations that RANEY® cobalt can facilitate. Hopefully this review will help redress matters in this regard. A second factor contributing to the lack of an extended “uptake” of the title reagent in chemical synthesis is the seemingly disparate array of techniques detailed in the literature for preparing RANEY® cobalt and its “doped”/ “spiked” variants, all of which appear to display quite distinct reactivity profiles. The discontinuation of the supply of certain previously commercially available sources of RANEY® cobalt87 has certainly exacerbated matters. Furthermore, the seemingly frequent need to use very high pressures of dihydrogen (and, often simultaneously, high temperatures) is also off-putting with the end result that even for all the potential chemoselectivities “on offer” the operational requirements associated with using the reagent just seem too daunting. We believe the drawbacks mentioned above are all readily addressed by using the simple protocols defined below17 that allow for the ready preparation of RANEY® cobalt possessing consistent and effective properties. Furthermore, by using larger quantities of RANEY® cobalt (e.g. ca. 200% w/w wrt substrate)34,79b in the reduction reactions many of these can be carried out using dihydrogen at atmospheric pressure and usually at room temperature. Our recent88 application of such protocols to the tandem reductive cyclisation of the enone 63 (Scheme 20) to compound 6489 (and embodying the ABCD framework of the important alkaloid vindoline90) has served to reinforce our enthusiasm for RANEY® cobalt as a catalyst for the chemoselective reduction of polyfunctionalised substrates. We contend that RANEY® cobalt is an under utilised but particularly effective catalyst for the chemoselective reduction of polyfunctionalised organic substrates, notably those incorporating nitro and nitrile moieties. As such, continued and systematic investigations into its catalytic properties are to be
7442 | Org. Biomol. Chem., 2014, 12, 7433–7444
Scheme 20
encouraged as these are very likely to reveal new and valuable patterns of reactivity.
Notes and references 1 (a) H.-U. Blaser, C. Malan, B. Pugin, F. Spindler, H. Steiner and M. Studer, Adv. Synth. Catal., 2003, 345, 103; (b) J. G. de Vries and C. J. Elsevier, Handbook of Homogeneous Hydrogenation, Wiley-VCH, Weinheim, 2007; (c) S. Nishimura, Handbook of Heterogeneous Catalytic Hydrogenation for Organic Synthesis, John Wiley and Sons, New York, 2001. 2 The development of metal-free hydrogenation systems is an emerging area that is likely to assume considerable significance in due course. For a recent review see: L. J. Hounjet and D. W. Stephan, Org. Process Res. Dev., 2014, 18, 385. 3 For useful historical introductions see: (a) M. Raney, Ind. Eng. Chem., 1940, 32, 1199; (b) D. S. Tarbell and A. T. Tarbell, J. Chem. Educ., 1977, 54, 26. 4 M. Raney, U.S. Pat., 1,628,190, 1927. 5 M. Raney, U.S. Pat., 1,915,473, 1933. 6 See the Johnson Matthey website for information on commercially available sponge metal catalysts: http://www.jmcatalysts.com/pct/marketshome.asp?marketid=241 (accessed March 2014). 7 A. J. Smith, L. O. Garciano II, T. Tran and M. S. Wainwright, Ind. Eng. Chem. Res., 2008, 47, 1409. 8 H. Adkins, Reactions of Hydrogen with Organic Compounds Over Copper-Chromium Oxide and Nickel Catalysts, The University of Wisconsin Press, Madison, 1937. 9 (a) H. Adkins and A. A. Pavlic, J. Am. Chem. Soc., 1947, 69, 3039; (b) H. Adkins and H. R. Billica, J. Am. Chem. Soc., 1948, 70, 695; (c) H. Adkins and H. R. Billica, J. Am. Chem. Soc., 1948, 70, 3118; (d) H. Adkins and H. R. Billica, J. Am. Chem. Soc., 1948, 70, 3121 See, also: (e) R. Mozingo, Org. Synth., 1941, 21, 15; (f) H. R. Billica and H. Adkins, Org. Synth., 1949, 29, 24; (g) A. I. Vogel, A Text Book of Practical Organic Chemistry, Longmans, London, 3rd edn, 1956, pp. 870–872. 10 T.-K. Yang and D.-S. Lee, Raney Nickel, in Handbook of Reagents for Organic Synthesis: Oxidizing and Reducing Agents, ed. S. D. Burke and R. L. Danheiser, John Wiley and Sons, Chichester, UK, 1999. 11 See ref. 1c, pp. 24–25. 12 W. Reeve and W. M. Eareckson III, J. Am. Chem. Soc., 1950, 72, 3299.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 17 June 2014. Downloaded on 07/10/2014 13:26:05.
Organic & Biomolecular Chemistry
13 A. J. Chadwell Jr. and H. A. Smith, J. Phys. Chem., 1956, 60, 1339. 14 The nature of the alkali-leaching process used to remove Al from the Co–Al alloy and its impact on the properties of the RANEY® cobalt so-formed has been the subject of a number of studies: (a) J. P. Orchard, A. D. Tomsett, M. S. Wainwright and D. J. Young, J. Catal., 1983, 84, 189; (b) S. Nishimura, M. Kawashima, S. Inoue, S. Takeoka, M. Shimizu and Y. Takagi, Appl. Catal., 1991, 76, 19; (c) K. Abaciouğlu and Y. Salt, React. Kinet., Mech. Catal., 2010, 101, 163. 15 B. V. Aller, J. Appl. Chem., 1957, 7, 130. 16 G. M. Badger, M. Kowanko and W. H. F. Sasse, J. Chem. Soc., 1959, 440. 17 Nickel-free cobalt–aluminium alloy (12.0 g of 3 : 7 w/w material provided as a
